1.
Navigator: This involves navigating Mantis spaces, gaining insights, reporting bugs, becoming super-users of the system, suggesting improvements and new features.
2.
Cartographer: This involves finding awesome datasets online, creating new spaces from CSV spreadsheets, literature searches, PDF documents, and demonstrating their usefulness.
3.
Data Scrapping: This involves creating new datasets for cartography, by scraping data, websites, Wikipedia, YouTube, Twitter, News, etc. Some programming and data science background is helpful, including using Selenium, BeautifulSoup, Scrapy, Octoparse, and other tools.
4.
Data Cleanup: This involves cleaning up scraped datasets for Mantis ingestion, which might need coding, macros, logical thinking, and semi-automated semi-manual cleanup, to make sure datasets are meaningful, helpful, and ready for cartography.
5.
Platform (Horizontal) Full-Stack Engineer: This involves the most coding and full-stack software development across front-/back-end, GitHub, integration, and is typically reserved to those with the most software engineering experience.
6.
Application (Vertical) Developer: This involves domain-specific coding, deep understanding of specific application areas, creating domain-specific multi-modal embeddings, solving domain-specific ML problems, and leading applications of Mantis to specific verticals
7.
Self-Contained Coding/ML Tasks: This involves Mantis-inspired coding of specific ML capabilities related to Mantis, which can then be integrated and plugged in, by working closely with platform developers.
Potential paths for new Mantis team members could be:
(a) 1,2,4,3 for those with less programming background, or
(b) 5,6,7 for those with extensive full-stack software engineering background in large codebases and teams
(c) 7,6,5 for those with more isolated coding and ML interests but less extensive software engineering